Machine Learning-Based Analysis of Magnetic Resonance Radiomics for the Classification of Gliosarcoma and Glioblastoma.
Machine Learning-Based Analysis of Magnetic Resonance Radiomics for the Classification of Gliosarcoma and Glioblastoma.
复制标题
基于机器学习的磁共振放射组学分析用于胶质肉瘤和胶质母细胞瘤的分类
DOI：
10.3389/fonc.2021.699789
复制
发表时间：
2021
影响因子：
4.7
通讯作者：
Chen X
中科院分区：
文献类型：
--
作者：
Qian Z;Zhang L;Hu J;Chen S;Chen H;Shen H;Zheng F;Zang Y;Chen X
目的寻找最佳的机器学习方法，用于胶质肉瘤（GSM）和胶质母细胞瘤（GBM）的放射学鉴别。材料与方法回顾性分析83例经手术病理证实的GSM患者的MRI资料（58例男性，25例女性;平均年龄，50.5 ± 12.9岁;范围，16-77岁）和100例GBM患者（58名男性，42名女性;平均年龄，53.4 ± 14.1岁;范围，12-77岁），并将他们随机分为训练和验证组。放射组学特征从肿瘤肿块和瘤周水肿中提取。三种特征选择和分类方法在区分GSM和GBM方面的性能进行了评估：最小绝对收缩和选择算子（LASSO），救济和随机森林（RF）;和adaboost分类器（Ada），支持向量机（SVM）和RF;分别。分析各方法的受试者工作特征曲线下面积（AUC）和准确度（ACC）。结果基于肿瘤质量特征的选择方法LASSO +分类器SVM在验证集中具有最高的AUC（0.85）和ACC（0.77），其次是Relief + RF（AUC = 0.84，ACC = 0.72）和LASSO + RF（AUC = 0.82，ACC = 0.75）。基于瘤周水肿特征，发现Relief + SVM在验证集中具有最高AUC（0.78）和ACC（0.73）。无论哪种方法，在区分GSM和GBM时，肿瘤肿块特征显著优于瘤周水肿特征（P < 0.05）。此外，最佳放射组学模型的灵敏度，特异性和准确性上级神经放射学家获得的结果。结论本放射组学研究确定了选择方法LASSO结合分类器SVM是基于肿瘤质量特征区分GSM和GBM的最佳方法。
Objective To identify optimal machine-learning methods for the radiomics-based differentiation of gliosarcoma (GSM) from glioblastoma (GBM). Materials and Methods This retrospective study analyzed cerebral magnetic resonance imaging (MRI) data of 83 patients with pathologically diagnosed GSM (58 men, 25 women; mean age, 50.5 ± 12.9 years; range, 16-77 years) and 100 patients with GBM (58 men, 42 women; mean age, 53.4 ± 14.1 years; range, 12-77 years) and divided them into a training and validation set randomly. Radiomics features were extracted from the tumor mass and peritumoral edema. Three feature selection and classification methods were evaluated in terms of their performance in distinguishing GSM and GBM: the least absolute shrinkage and selection operator (LASSO), Relief, and Random Forest (RF); and adaboost classifier (Ada), support vector machine (SVM), and RF; respectively. The area under the receiver operating characteristic curve (AUC) and accuracy (ACC) of each method were analyzed. Results Based on tumor mass features, the selection method LASSO + classifier SVM was found to feature the highest AUC (0.85) and ACC (0.77) in the validation set, followed by Relief + RF (AUC = 0.84, ACC = 0.72) and LASSO + RF (AUC = 0.82, ACC = 0.75). Based on peritumoral edema features, Relief + SVM was found to have the highest AUC (0.78) and ACC (0.73) in the validation set. Regardless of the method, tumor mass features significantly outperformed peritumoral edema features in the differentiation of GSM from GBM (P < 0.05). Furthermore, the sensitivity, specificity, and accuracy of the best radiomics model were superior to those obtained by the neuroradiologists. Conclusion Our radiomics study identified the selection method LASSO combined with the classifier SVM as the optimal method for differentiating GSM from GBM based on tumor mass features.
登录
查看更多内容
影响因子：
3.9
作者：
Castelli, J.;Feuvret, L.;Noel, G.
通讯作者：
Noel, G.
影响因子：
5.9
作者：
Suh, Hie Bum;Choi, Yoon Seong;Lee, Seung-Koo
通讯作者：
Lee, Seung-Koo
MRI 特征通过机器学习方法预测低级别神经胶质瘤中的 p53 状态。
DOI：
10.1016/j.nicl.2017.10.030
发表时间：
2018
期刊：
NeuroImage. Clinical
影响因子：
--
作者：
Li Y;Qian Z;Xu K;Wang K;Fan X;Li S;Jiang T;Liu X;Wang Y
通讯作者：
Wang Y
DOI：
10.18632/aging.101594
发表时间：
2018-10-22
期刊：
Aging
影响因子：
--
作者：
Qian Z;Li Y;Sun Z;Fan X;Xu K;Wang K;Li S;Zhang Z;Jiang T;Liu X;Wang Y
通讯作者：
Wang Y
影响因子：
1.3
作者：
Wang S;Zhou J
通讯作者：
Zhou J